The Death of High Fidelity

Rock and pop producers have always used compression to balance the sounds of different instruments and to make music sound more exciting, and radio stations apply compression for technical reasons. In the days of vinyl rec- ords, there was a physical limit to how high the bass levels could go before the needle skipped a groove. CDs can handle higher levels of loudness, although they, too, have a limit that engineers call "digital zero dB," above which sounds begin to distort. Pop albums rarely got close to the zero-dB mark until the mid-1990s, when digital compressors and limiters, which cut off the peaks of sound waves, made it easier to manipulate loudness levels. Intensely compressed albums like Oasis' 1995 (What's the Story) Morning Glory? set a new bar for loudness; the songs were well-suited for bars, cars and other noisy environments. "In the Seventies and Eighties, you were expected to pay attention," says Matt Serletic, the former chief executive of Virgin Records USA, who also produced albums by Matchbox Twenty and Collective Soul. "Modern music should be able to get your attention." Adds Rob Cavallo, who produced Green Day's American Idiot and My Chemical Romance's The Black Parade, "It's a style that started post-grunge, to get that intensity. The idea was to slam someone's face against the wall. You can set your CD to stun."

It's not just new music that's too loud. Many remastered recordings suffer the same problem as engineers apply compression to bring them into line with modern tastes. The new Led Zeppelin collection, Mothership, is louder than the band's original albums, and Bendeth, who mixed Elvis Presley's 30 #1 Hits, says that the album was mastered too loud for his taste. "A lot of audiophiles hate that record," he says, "but people can play it in the car and it's competitive with the new Foo Fighters record."

Just as cds supplanted vinyl and cassettes, MP3 and other digital-music formats are quickly replacing CDs as the most popular way to listen to music. That means more conven- ience but worse sound. To create an MP3, a computer samples the music on a CD and compresses it into a smaller file by excluding the musical information that the human ear is less likely to notice. Much of the information left out is at the very high and low ends, which is why some MP3s sound flat. Cavallo says that MP3s don't reproduce reverb well, and the lack of high-end detail makes them sound brittle. Without enough low end, he says, "you don't get the punch anymore. It decreases the punch of the kick drum and how the speaker gets pushed when the guitarist plays a power chord."

But not all digital-music files are created equal. Levitin says that most people find MP3s ripped at a rate above 224 kbps virtually indistinguishable from CDs. (iTunes sells music as either 128 or 256 kbps AAC files — AAC is slightly superior to MP3 at an equivalent bit rate. Amazon sells MP3s at 256 kbps.) Still, "it's like going to the Louvre and instead of the Mona Lisa there's a 10-megapixel image of it," he says. "I always want to listen to music the way the artists wanted me to hear it. I wouldn't look at a Kandinsky painting with sunglasses on."